Freshwater Biology (1988) 19, 95-108
Aspects of mating, reproduction, and co-occurrence
in three freshwater calanoid copepods
PATRICIA CHOW-FRASER* and EDWARD J. MALY Concordia University,
Department of Biology, Montreal, Quebec, Canada
SUMMARY. 1. We examined the seasonal distributions and mating
behaviour of Epischura lacustris, Diaptomus minutus and Diaptomus
oregonensis to evaluate several proposed mechanisms for reproductive
isolation. There was no divergence in onset of maturity or timing of
breeding periods between sympatric diaptomid populations; size displacement between co-occurring diaptomid species did not result in concomitant reduction in mating frequency between heterospecific pairs in the
laboratory.
2. Divergence in mating behaviour was supported by observations that,
in conspecific mating experiments, males of both diaptomid species discriminated between females bearing ripened ovaries (gravid) and those
that did not (non-gravid), and that, in heterospecific trials, males either
mated indiscriminantly with all females, or engaged in mating very
infrequently.
3. In mating experiments with Epischura. males did not discriminate
between gravid and non-gravid females, but mated exclusively with virgins
of a specific relative size. Divergence in timing of breeding seasons
between Epischura and the two diaptomid species suggests that even if
mating between genera is possible, it does not occur often in nature.
4. We discuss the adaptive significance of different reproductive strategies of these three copepods and speculate on mechanisms that allow for
coexistence of the closely related taxa in nature.
Introduction
Reproductive isolation is prerequisite to maintenance of taxonomically distinct populations
when closely related species co-occur in nature.
Not surprisingly, a great deal of research has
been conducted to uncover mechanisms that
*Presenl address: University of Alberta, Department of Zoology. Edmonton. Alberta. Canada TiSG
2E9.
Correspondence: Dr E, J. Maly. Department of
Biology, Concordia University. 1455 de Mais<inneuve
Blvd- W . Montreal. Quebec. Canada H3G IMS.
lead to reproductive isolation. But while mechanisms are often proposed, the task of validating
such hypotheses is slow to follow suit. In this
study, we try to remedy this situation by using
information on the reproductive biology of three
calanoid copepods to evaluate applicability of
several proposed mechanisms.
The copepods studied include Epischura lacuslria Forbes, Diaptomus minulus Lilljeborg.
and D. oregonensis Lillieborg. These copepods
frequently assort in pairs iti lakes of south-eastern Quebec (Maly. 1984), and are three of the
95
96
Patricia Chow-Fraser and Edward J. Maly
most commonly occurring crustaceans in glaciated eastern North America (Carter et al.,
1980). They share a similar ecology in lakes
(Sandercock, 1967; Rigler & Langford. 1967.
being mainly herbivorous in many of their lifestages (Chow-Fraser, 1986; Chow-Fraser &
Wong, 1986). The two diaptomids are similar in
size, and when they are found sympatrically.
sometimes exhibit size displacement (Maly.
1984) or partitionitig in their temporal or spatial
distributions (Sandercock, 1967; Rigler & Langford. 1967).
Divergence in size and/or distribution has
sometimes been viewed as evidence that interspecific competition has been a selective pressure (Hutchinson. 1951; Cole. 1961; Maly.
1984). In this study we consider an alternative
view, that size displacement has evolved to prevent interbreeding between very similar-sized
species. As such, it is a form of morphological
divergence that has already been proposed as a
mechanism for reproductive isolation among
calanoid copepods (Davis. 1961; Fleminger.
1967. 1975; Blades & Youngbluth. 1980; Jacoby
& Youngbiuth. 1983).
To evaluate these proposed mechanisms, we
conducted laboratory mating experiments with
lake-resident populations of the three copepod
taxa. The factors we considered include relative
size of mating pairs, reproductive condition of
females (gravid versus non-gravid), and taxonomic affiliation of each mating partner. In addition, we compared seasonal distributions of allopatric and sympatric populations and relate
thesefindingsto differences in reproductive and
life history characteristics.
Another mechanism that may apply to caianoid copepods is divergence in behaviour, in
addition to. or instead of morphology. This is
based on the observations that: (1) many calanoid copepods display species-specific mating
rituals (Blades & Youngbluth, 198U; Jacoby &
Youngbluth. 1983), and (2) that several diaptomid species selectively engage in pre-copulatory coupling with gravid versus non-gravid
females (Watras. 1983). Since in Diaptomus
only gravid females produce viable clutches
when mated (Watras. 1983). these observations
suggest that males have not only evolved to
recognize females of their own species, but also
to discriminate reproductively receptive individuals from those that are not. Such discriminatory
power may effectively serve to prevent hybridization between species, as long as males do not
mate indiscriminantly with gravid females of any
species.
Field collection
A third mechanism that may apply is divergence in timing of reproductive events. This
hypothesis maintains that species that do not
exhibit divergence in behaviour or morphology,
may still experience reproductive isolatioti if
there is temporal partitioning in their breeding
periods (Davis, 1961; Sandercock. 1967; Maly.
1976; Jacoby & Youngbluth. 1983).
Methods
We selected four lakes, Trouser, Nick. Sally and
Moulin Lakes, in which the three copepod taxa
were found in various combinations. Nick and
Trouser Lakes contained all three. Sally contained Epischura and D. minutus, and Moulin
contained only D. oregonensis. These lakes are
located in the vicinity of Montreal and their
physico-chemical characteristics have been published elsewhere (Maly. 1984).
Quantitative zooplankton samples (2-litre
Schindler-Patalas traps; mesh size 110 //m)were
collected from top and bottom strata at several
locations iti each lake. Ten replicate samples
were taken from each lake and were preserved in
the field with 4% formalin solution. In total, six
sampling trips were made from the latter part of
May to the end of August in 1986. In the laboratory, adult stages of each copepod taxon were
identified and enumerated with the aid of a
dissecting microscope.
Qualitative samples were also collected in
diagonal net tows (mesh size 80 /^m) and preserved with formalin solution in the field. In the
laboratory, an aliquot of each sample was withdrawn and examined under a dissecting microscope. Adult and copepodid stages of each diaptomid species were identified and enumerated.
Whenever possible, at least 100 individuals of
each taxon were counted in each sample. From
this, percentage composition data were calculated for copepod life-stages.
Mating experiments
Copepods were collected in diagonal net tows
from each of the four lakes at various times
Mating in three freshwater calanoid copepods
during the ice-free season. They were kept in
lakewater in 30-litre plastic carboys and transported back to the laboratory, usually within 4—
6 h of collection. Animals from each lake were
transferred to separate aerated glass aquaria.
kept at between 20 and H^C, and subjected to a
natural photoperiod. Copepods were fed aliquots of an algal suspension of Chlorella sp.,
Chlamydomona.s sp., Ankistrodesmus sp.. Seenedesmtis sp.. and unidentilied microflagellates
that had been cultured in a goldfish aquarium.
The algal mixture was added immediately after
animals were placed in aquaria, and daily thereafter until experiments were terminated. We
only used animals that were actively swimming
and that had been kept in captivity for less than 3
days. Animals from the maintenance aquaria
were isolated in depression plates with the aid of
a dissecting microscope. One male and one nonegg-bearing female with no attached spermatophores were then placed into a 1000 ml
holding beaker and then into a clear plastic vial
containing approximately 30 ml of 64-^m-filtered lakewater. The source of lakewater was
the lake from which animals were obtained;
when animals from two different lakes were
involved, lakewater from one lake was arbitrarily employed. The reproductive condition
of the female was noted as gravid or non-gravid;
a female was considered gravid when most of the
oocytes had acquired a dark coloration. Vials
were kept under the same conditions as maintenance aquaria (but not aerated). After a period
of from 3 to 24 h, copepods were killed and
preserved with the addition of 10 drops of 37%
formalin solution. Whenever possible, experiments with five to ten replicate pairs were performed simultaneously. Depending on availability of animals in the maintenance aquaria.
isolation and sorting consumed from I to 4 h for
every ten replicates run. Pairs of animals from
the same and different lakes were included in
these experiments. In total. 744 experiments
were conducted (see Table 1).
Presence of a spermatophore or spermatophores on female genital pore or urosome,
male antennule. or bottom of the vial was considered evidence for a mating attempt. This
method of scoring does not take into consideration attempts that might have been made but
that did not result in extrusion of spermatophorts, nor does it distinguish between
single and multiple transfers. No effort was
made to determine whether or not the spermatophore was full or empty. We noted whether
or not females were bearing eggs at the end of
the experiment, and measured the prosome
length of copepods to calculate a size-ratio
{female/male prosome length).
In a series of experiments we determined the
effect of experimental duration on mating frequency. Results from 3-h and 24-h experiments
were statistically compared with Chi-square
analysis. We found no evidence that length of
experiment had any effect on the data
(X'-O.IO6; fXI.75). Secondly, by monitoring
pairs of copepods in 1 ml micropipette plates,
we confirmed that mating, if attempted at all,
took place within the first 2h of coupling. Thus,
duration of the mating experiment was not an
important consideration here, since duration of
al! experiments exceeded 2h.
Although Epi.schtira were present in Nick and
Trouser Lake, they were so rarely found that
mating experiments were only performed with
animals from Sally Lake. For the same reason,
very few experiments were conducted with D.
minuius from Nick Lake. There was also a disproportionate representation of both ailopatric
diaptomid populations in the dataset; sympatric
populations were generally found in low numbers except in July.
TABLE 1. Summary scores of mating experiments, DM^D.
NS = not sigiiificanl (P>0.05).
Female
species
DM
X>O
PO
DM
DM
DO
DM
DO
minutus; DO = D. oregonensis;
Not gravid
Gravid
Male
species
97
Mated
Not
mated
Mated
Not
mated
Chi-square
P
18
25
7
1
20
27
32
23
14
25
30
4
125
123
173
97
27.57
19.96
0.14
0.00
0.001
0.001
NS
NS
98
Patricia Chow-Fraser and Edward J. Maly
We analysed mating frequency data by means
of appropriate Chi-square tests (Sokal & Rohlf,
1981). To test for mutual independenee of size
ratio, reproductive condition and frequency of
mating attempts; we used a three-dimensional
contingency table (Zar, 1984).
Life history and reproductive characteristics
Copepods were captured and maintained as in
mating experiments. As soon as it was convenient (usually within 6 h of collection), egg-bearing females were isolated and placed individually in plastic vials containing 30 ml of filtered
lakewater. Vials were also kept under the same
conditions as in maintenance aquaria. We recorded whether or not females were gravid, and
then determined their clutch size. Condition of
the females was monitored at approximately the
same time each day until they expired. We recorded whether she still carried eggs and/or was
gravid, and whether or not nauplii were visible if
she no longer carried eggs. Nauplii, if present,
were subsequently removed. Water in the vials
was changed approximately every 5 days, and
approximately 1ml of algal mixture (as above)
was added two or three times per week, We
measured prosome length of the female upon
her death and discarded contents of the vial.
These experiments were replicated twice (once
each in July and August) for Nick and Trouser
animals and three times (one each in June. July
and August) for Moulin and Sally Lakes.
To determine the period between final moult
and initial gravid condition, we isolated thirty to
fifty late copepodid stages (C3 or C4) and placed
them individually into experimental vials as
above. They were also held under the same
100
Trouser
50
100
50
\
100
Solly
\
50
\
100
Moulin
'
50 -
*
—
May
June
July
' D oregonensis
• D minutus
• —• Epischura
August
FIGS. 1. Relative abundance of adult calanoid copepods plotted against sampling date. Values are percentage of
the maximum observed for each species.
Mating in three freshwater calanoid copepods
conditions as for egg-bearing females. Animals
were monitored daily for production of moults.
Upon their final moult, females were scored
daily as gravid or non-gravid. For males, only
time of death was noted. Owing to the lengthy
duration of these experiments, they were not
repeated.
Results
99
were similarly observed in July. Thus, females of
both diaptomid species matured in large numbers in July, regardless of presence of congeneric
species.
By contrast, adult Epischura exhibited their
seasonal maxima in May and June, and quickly
disappeared in July: by mid-August a second
generation of adults appeared but was usually
found in low numbers. This pattern of abundance was noted regardless of co-occurence of
Diaptomu.'i.
Seasonal distribution
In Nick and Trouser Lakes, where the diaptomid species occur sympatrically. seasonal
maxima for both species of adults occurred during July (Fig. 1). In Sally and Moulin Lakes,
where they occur allopatricaly, seasonal maxima
Examination of percentage composition for
the various diaptomid populations partially confirmed the trend noted above (Fig. 2). There was
no divergence in onset of maturity between sympatric diaptomid populations. Peaks in the relative abundance of adult diaptomid species in
Trouser
60
40
DM
20
0
Nick
60
40
g 20
E
0
-^
60
,..*•,.
Solly
40
20
0
Moulin
FIG. 2. Percentage of the total copcpodid iirul jdult popLilation thiit was ddull ftmalc in samples plotted against
sampling date. D. minutus in Nick Lake were present in negligible numbers throughout the summer.
100
Patricia Chow-Fraser and Edward J. Maly
Trauser Lake overlapped in mid-July; there
were unfortunately insufficient data to substantiate this trend in Nick Lake, By comparison.
there were no peaks in percentage composition
data corresponding to the ailopatric populations. In both Sally and Moulin Lake, percentage composition of adults in samples did not
decline during the summer after the initial increase in mid-June.
Conspecific mating between diaptomids
Experiments were conducted with animals
collected from all four lakes over a 3-month
period. Before we pooled data by species (to
increase sample size for statistical rigour), we
first investigated whether or not frequency of
mating attempted was affected by time of year.
With some variation, approximately the same
proportion of animals attempted to mate in each
month, proportions usually ranging from 2H9( to
33% for D. oregonensis (Fig, 3a) and from 23%
to 26% for D. minutus (Fig. 3b). Effect of lake
origin was also minor, since mating frequency
corresponding to different lakes were very
similar.
When low mating frequencies were recorded,
corresponding proportion of gravid animals in
samples were also low (Fig. 3c and d), suggesting
that effect of reproductive condition superceded
any effect of season or lake origin on frequency
of mating attempts. Since this analysis demonstrated that time of year had minimal effect on
mating data, we proceeded to pool data by
species, and test the effect of reproductive condition and size-ratio on frequency of mating
attempts.
The three-dimensional contingency table, appropriate for testing mutual independence of
three variables (Zar, 1984) was used to examine
the simultaneous effects of sex-size-ratio and
reproductive condition on mating data. The data
were pooled by species, regardless of lake
origin. In both cases the null hypothesis of mutual independence was rejected {^=lK).yi.,
df-34. /'<0,001 for D, minutus; x--57,85.
60 - ( a ) D oregonensls
60 - { c) 0 oregonensis
40
40
20
May
June
July
Aug.
60 - ia) O minutus
0
20
40
60
40
60
60 h ( d 1 D minutus
40
40
20
20
May
June
July
'^o"fh
Aug.
0
20
Gravid 9 (%)
FIG. 3. (a) Percentage of males that attempted lo mate plotted against experiment date for Diapiomiis orei-nnensis
data, (b) Same as {;i) but for D. minuiiis diita. (c) Percentage of males thai attetnpted to mate plotted as a function
of per cent gravid females iti sample for D. arcgoncnsi.s data, (d) Same us (c) but for D. mimtius data.
Mating in three freshwater calanoid copepods
df=40, P<().05 for D. oregonensis). We then
proceeded to determine the effects of each variable separately.
We tested to see if reproductive condition
alone could influence mating frequency. For
both species of Diaptomus. without taking sizeratio into consideration, a statistically higher
proportion of males attempted to mate with
gravid than with non-gravid females (Table 1),
thus confirming the strong effect of reproductive
condition on mating attempts noted earlier (Fig,
3c and d).
Next, we tested if size-ratio alone could influ-
101
ence mating frequency. When data were pooled
by species, regardless of lake origin and reproductive condition, we found no significant effect
of sex-size-ratio on frequency of mating attempts for either species; however, once we
accounted for reproductive condition, we found
a significant effect for gravid individuals of D.
minutus (Table 2, Fig, 4), By contrast, size-ratio
did not significantly affect mating frequency of
D. oregonensis males although somewhat fewer
attempts were recorded at size ratios near unity,
at least over the range of ratios examined (Fig,
4).
TABLE 2. Summary statistics from tests of independence of
mating frequency on size-ratio. DM — D. minulus:
DO-D.
oregonensis: G = gravid; NG = non-gravid; NS^non-signifieant
(P>0.05).
Reproductive
condition
n
DF
Chi-square
P
DM
G + NG
G
NG
169
38
131
5
5
5
8.74
12.08
4.41
NS
0,05
NS
DO
G + NG
G
NG
20()
52
148
6
6
6
5.56
1.28
10.18
NS
NS
NS
Species
D. minutus
Non-qrovid ?
D. oregonensis
Non-gravid ?
0-8
0-6
n
0-4 15
5
5
0-2
GrQvid
0-6
0-4
0-2
0
3 2 2
1-00
1-12
1-24
1-00
1-12
•24
Sen-size rotio
FIG. 4, Frequcney of mating attempts plotted as a funclion ol corresponding size-ratio. Numbers in the body of the
figure indicate total number of experiments used to calculate frequencies.
102
Patricia Chow-Fraser and Edward J. Matv
Heterospecific mating between diaptomids
Mating in Epischura
We first tested results of heterospecific mating
experiments for mutual independence of reproductive condition and frequency of mating attempts (Table 1). The proportion of males that
attempted to mate with gravid females did not
differ significantly from that with non-gravid
females, indicating that gravid condition was not
an important cue for heterospecific coupling. In
general, frequency of mating attempts was low
compared with those of conspecific mating experiments (Table 1). especially when D. minuttts
males were paired with D. oregonensis females.
The reproductive biology of Epischura is fundamentally different from that of Diaptomus.
First, unlike Diaptomus, which usually detaches
spent spermatophores before extrusion of a
clutch, Epischura retains it for a considerable
period, even while a clutch is being extruded.
Secondly, whereas Diaptomus extrudes the eggs
into a sac. bears the clutch for 1-3 days, and then
sheds it as a unit. Epischura extrudes eggs individually, and releases them immediately without
bearing the clutch in a sac. If eggs are borne in a
clutch, the clutch breaks apart relatively quickly.
This has the effect of scattering eggs over a
considerable distance, especially if the female is
swimming during the process. The third major
difference is that diaptomid eggs are brooded in
the egg sac, whereas eggs of Epi.schura normally
develop in the water column or sediment, away
from the female. Egg development in Epischura
requires 3 or 4 days at lifC in the laboratory.
We grouped the pooled data for heterospecific
mating between D. oregonensis males and D.
minutus females by size-ratio, and plotted relative frequency of attempts against respective
ratios (Fig. 5). Despite the overall negative relationship between variables (r--(1.14). the large
scatter in data rendered the relationship statistically non-significant (P>().5). Even when we
sorted the data by lake, we found no significant
relationship between relative frequency and
sex-size-ratio for sympatric populations.
TABLE 3. Mean clutch size of Epischura females.
Clulch order refers to order in which multiple clutches
were extruded. ^i =
With few exceptions, heterospecific pairing of
D. mintitits males with D. oregonensis females
did not result in spermatophore transfer; only
iive of 123 males attempted to mate. There was
also no significant effect of reproductive condition or size-ratio on the mating data. These
observations indicate that D. minutus ma\es ciinnot or do not engage in heterospecitic copulation
to the point where a spermatophore is
transferred.
Month
June
August
Clulch
order
Clutch size
n
SE
First
Second
Third
Fourth
First
Second
16.21
10. S4
5.55
6.(H)
24.07
11.40
33
19
11
2
0.76
1.41
0.99
-
14
5
!.H2
2.(11
Q oregonensis a x 0 minutus ?
Mating frepuisncy
- SympatrJc
0-8
5
12
22
• 22
0-6
24
0'4
6
7
23
10
17 23
0-2
1
*
0-98
14
20 3
7
•
1
1
1
\
1
1
•
0-82
Sen-size ratio
•
6 4 2
•
f
1
0-70
FIG. 5. Frequency of mating altempts plolled as a funclion of corresponding size-ralio. Arrows indicale ihe range
in ralios ihat correspond to respective experiments mvolving allopairic and sympatric populations. Numbers in the
body of ihe iigure as in Fig. 4.
.Mating in three freshwater calanoid copepods
Lastly, Diaptomus only produces one clutch per
mating; Epischura. on the other hand, can produce up to four clutches from one spermatophore attachment (Table 3),
Given that female Epischura can produce
multiple clutches from a single mating, two new
categories were examined: females that had previously mated, and those that had not (virgins).
Females in this latter category were newly
moulted females 1-2 days old. Those in the first
category included females that had retained
their spent spermatophores and a few whose
spermatophores had become dislodged from
their genital pores.
Males in these experiments only attempted to
mate with virgin females (Fig, 6); however,
ovary development in these virgins did not have
any significant effect on mating frequency.
There was a significant effect of size-ratio
{X-=H.59, df-6; /'<0.025) since mating attempts were recorded exclusively for experiments with corresponding sex-size-ratios between 1.08 and 1.14, Because relative frequencies were all 1.0, it was not possible to determine
peak frequency.
Life history and reproductive characteristics
The mean life span of D. minutus females
collected from Trouser and Sally Lakes were
39,0 and 41,7 days, respectively (Fig. 7a), These
values are similar to those of D. oregonensis
from Trouser, Nick and Moulin Lakes (30,6,
34,0 and 44,4 days, respectively), A two-way
analysis of variance (based on unequal sample
size) revealed no statistical differences in mean
female lifespan from lake to lake, or from species to species (/'>0,05), The lifespan of diaptomid males was generally less than that of
females within the same lake. There were insufficient data from Nick and Trouser Lakes to
facilitate statistical analysis, but an analysis of
variance for remaining data indicated that, regardless of species, mean lifespan of males did
not differ among lakes (/*>0,05).
Proportion of females" adulthood spent in the
gravid condition was computed for both species.
D. /Timu/Mi females spent 55-60% of their adulthood in the gravid condition, whereas D.
oregonensis only spent 20-30% of their adulthood in this condition (Fig, 7b). The interval
between gravid periods for D. minutus was also
significantly reduced compared with that for D.
oregonensis (F<0,05), even though there was
great among-iake variation within species (Fig,
7c),
The number of days between final moult and
the initial gravid period also differed significantly between species (Fig, 7d) under our controlled laboratory conditions; that for D. minutus ranged from 2 to 3 days, while that for D,
oregonensis ranged from 4,5 to 6 days. The mean
Virgins
10 -
Moted
Gravid 2
Grawid
Jl
Non-gravid ?
Non-grovid 9
• Jm
•06
1-16
103
1-24
!^08
•16
Sex-size fatio
FIG. 6. Frequency of mating attempts for corresponding size-ratios.
•24
104
Patricia Chow-Fraser and Edward J. Maly
D
minutus
D
oregonensis
U
rninutus
HD
or egonen
*
(d)
40
£
II
20
(D)
t
O
1-
_,
-
t
60
(e]
h
6 -
40
2 -
(tl
(c)
*
*
-
n
PI
r1
n
o
0
1 f
FfG, 7. Summary of reproductive and life history characteristics for diaptomid species: (a) mean± 1 SE adulthood
of males (striped) and females (open); (b) niean±l SE proportion of female adulthood spcnl in gravid condition;
(c) mean±i SE interval between gravid periods; (d) mcan±l SE interval between fmal moult and initial gravids
period; (e)mean±l SE duration of gravid period; (Oniean±i SE duration beiwcen clutch formation and onset of
gravid condition (duration in days). T^Tiouser Lake; N^Nick Lake; S-Sally Lake; M = Moulin Lake.
duralion of the gravid period for D. minutus
females was more than 3 times higher than that
for D, aregonensis females (Fig. 7e). We also
calculated the mean duration between clutch
formation (when the female carried one at the
start of the experiment) and onset of the next
gravid period. D. oregonensis fem3\esm general
required longer than 3.5 days to accomplish this,
whereas D. minutus required less than 2,5 days
(Fig, 7f).
Results of these experiments indicated that
reproductive characteristics of these two diaptomids are species-specific, and for the most part
are unaffected by lake origin. Since conditions in
these experiments were identical, differences
between species cannot be attributed to environmental factors such as food, temperature or
presence/absence of mates, unless species respond differently to these factors.
Copepods varied greatly with respect to clutch
size from population to population in the field
(Fig, 8), A two-way atialysis of variance showed
that lake origin and month of year had significant
effects on clutch size (/'<0.05). This is not unexpected since clutch size is directly related to
temperature and food conditions (Elmore,
1983), and these fluctuate greatly from May to
August in lakes, Ihe most notable trend is that
clutch size of D. oregonensis was generally larger
than that of D. minutus, except in August, ai
which time only animals in Trouser Lake bore
large clutches.
Discussion
We can now review relevant information to
evaluate applicability of some of the proposed
mechanisms for reproductive isolation. Since
there was no apparent temporal partitioning
(Fig. 1), no apparent divergence in onset of
maturity (Fig. 2), and no summer seasonality in
laboratory experiments with respect to mating
frequency (Fig, 3) for the two diaptomid species.
Mating in three freshwater calanoid copepods
105
D oregonensis
D minufus
i
12
20 i
23
24
-
A
20
i
4
18
3
20
FIG. 8 Summary of mean (±1 SE) clutch size of diaptomids. T=Trouser Lake; N = Nick Lake; S=Sally Lake;
M-Moulin Lake. V, June; c, July; •, August.
we can eliminate divergence in timing of reproductive events as a possible mechanism.
We also found no evidence that divergence in
size (Maly. 1984) prevented interbreeding between sympatric diaptomid populations. Incidence of heterospecific mating in organisms
taken from Trouser Lake was nearly as high as
that noted for organisms taken where they occur
allopatrically, even though corresponding sizeratios in the former were substantially larger
(Fig. 5). Thus, size displacement between sympatric populations was not accompanied by a
significant concomitant reduction in heterospecific mating.
By the process of elimination, the mechanism
that probably applies here is divergence in behaviour associated with development of speciesspecific cues for preferential detection of gravid
females or species-specific receptivity of gravid
females for mating partners. Several observations lend credence to this. The fact that males
mated discriminantly with conspecific gravid
females, but did not distinguish gravid from nongravid heterospecific females, implies that the
major attraction is species-specific (Table 1;
Figs. 4 and 5). We also know that cues responsible for this initial attraction are probably
chemosensory (Kittredge et al., 1974; Katona.
1973, 1975) rather than morphological in nature
since non-gravid females were usually ignored
even though gravid females of similar size were
mated. We therefore support speculations made
by many others that sex pheromone is the primary male attraction in mating of calanoid
copepods. We further speculate that production, release and detection of this pheromone is
species-specific, and thus functions as a mechanism for reproductive isolation in Diaptomus.
We cannot, however, preclude the possibility
that gravid females are also exhibiting pre-copulatory behaviours to attract males, which are
therefore more receptive to mating than nongravid females.
The advantage in short-circuiting a potentially
unsuccessful copulation at the 'attraction' phase
is that males do not waste gametes, and do not
spend unnecessary energy in pursuit of nonreceptive conspecific, or heterospecific females;
at the same time, females also benefit from this
since they do not waste their gametes (transfer of
heterospecific spermatophores results in extrusion of non-viable clutches (Jacoby & Youngbluth, 1983)).
There also appears to be an interactive effect
of reproductive condition and size-ratio on mating frequency of D. minutus (Fig. 4b); peak in
mating activity associated with size-ratio was
only observed when gravid females were consid-
106
Patricia Chow-Frascr and Edward J, Maly
ered. The ratio associated with this peak falls
within I.IU and 1.12, bracketing the observed
"optimum' size-ratio of 1.10 that corresponds to
peak mating intensity in field populations of D.
minuttis observed by DcFrenza et al. (1986).
Thus, data collected in these relatively simple
laboratory trials reflect behaviour exhibited by
natural diaptomid communities. By comparison, the interactive effect between reproductive condition and size-ratio for D. oregonensis was more difficult to ascertain since no distinct peak in mating frequency was evident (Fig.
4d), but the range of sex-size-ratios tested was
not as broad as that examined by DeFrenza et al,
(198ft).
Results from mating experiments with Epischura point to a somewhat different scenario
(Fig. 6). Since Epischura did not discriminate
between gravid and non-gravid females, release
of pheromones. if produced by this species, is
probably independent of ovary development,
and most likely commences soon after the final
moult. Correct placement of a spermatophore
on the female's genital pore may inhibit further
production and release of this pheromone
(Hopkins & Mackin. 1977), since males did not
attempt to mate with females that had spermatophores attached to them from a previous
mating. The fact that males did not copulate with
previously mated females, even when spermatophores were subsequently dislodged, also
implies that production and release of pheromone do not re-occur once the process has
been arrested.
Thai sex-size-ratio should have an effect on
frequency of mating attempts for D. minutus
(Fig. 4) and Epischura (Fig. ft) suggests that
morphological cues may also function in mate
recognition more strongly in these species than
in D. oregonensis, assuming that cues operate
after initial contact has been made. Epischttra
may have evolved mate recognition based on
relative size because there is virtually no other
large calanoid present in the lake to interfere
with successful copulation at the time when Epischura engages in peak mating activity (Fig. 1),
at least near Montreal. In such a situation, any
femaleof appropriate size may be mated with no
attending ill consequences. This behaviour may
also capitalize on the fact that females apparently only mate once (Fig. 6), so that males who
wait until females become gravid are disadvantaged.
The reason for D, minutus' apparent strong
reliance on morphological cues is more complicated to explain and may involve reproductive
and life history characteristics of both diaptomid
species. Adult females of hoth species have statistically similar survivorship when cultured in
the laboratory (Fig. 7a). Under controlled conditions, the main difference between species is
that females of D. minutus spend on the average
H)7c. of their adulthood in the gravid condition,
whereas D. oregonensis spend approximately
30% of their adulthood in this condition (Fig.
7b-f).
Given that gravid condition is prerequisite to
successful mating, the fact that D. minutus
females are gravid much more often than are D.
oregonensis implies that the former should be
more 'available' for mating. Since males of D,
minutus may be less constrained by availability
of receptive females, the added criterion they
use to select suitable mates, is an asset rather
than a liability, as supported by the low incidence of heterospecific mating between D, minutus males and D. oregonensis females. By contrast, D. oregonensis females are gravid only a
small proportion of their adulthood, so that
males must take full advantage of times when
females are gravid. By virtue of their poorer
discrimination, however, males engaged in heterospeciik mating with / ) . minutus females at a
higher frequency (Fig. 5), and thereby wasted
many more gametes.
The reproductive strategies for these caianoid
copepods appear to be a trade-off among the
following: (1) degree of discrimination required
for mate selection, (2) availability of mates, and
(3) clutch size. In D, minutus, clutch size of
females varies from two to seven eggs (Fig. 8),
and males appear to be extremely selective,
probably relying on both chemical and morphological cues for mate recognition, however,
females compensate for small clutch size by
being frequently available to males. In D.
oregnncnsis. females are less frequently available for mating, but males appear to be less
discriminating in choice of mates. Females, however, compensate for less frequent reproduction
by producing considerably larger clutches (6-13
eggs/ciutch. Fig. S). In Epischura, females appear to be available for an extremely short time
period, and males mate only with virgin females;
females, however, bear multiple clutches from a
single mating (6-23 eggs/clutch. Table 3).
Mating in three freshwater calanoid copepods
Whereas sex pheromones arc probably involved in distance location of females by males,
behavioural and morphological cues are probably more important for post-contact recognition.
This is supported by our observation that sizeratio had a significant effect on mating frequency
in mating trials involving gravid females (Fig. 4).
Future experiments should be conducted to determine if the sex pheromone is involved in mate
location and mate recognition.
Diaptomid species are often found sympatrically in many North American lakes and
ponds. The ability to copulate discriminantly
with reproductively compatible mates probably
helps to perpetuate a successful coexistence in
spite of their ecological similarity. We hope that
this study will further stimulate investigators to
identify the sex pheromone that may be responsible for mate attraction and recognition, and to
ascertain the species-spccilic nature of this
pheromone.
Acknowledgments
We wish to thank S. Plantc and D. Lacroix for
their capable technical assistance in the field.
and for their dedication to sorting copepods in
the laboratory. We also thank R. Kirner, A.
Elliott, J. DeFrenza, H. van Leeuwen and G.
Grad for their generous assistance. The
cooperation of cottagers and government agencies who permitted us access to study lakes is
gratefully acknowledged. The comments and
suggestions of an anonymous referee improved
this manuscript. This study was iinanced by the
Natural Sciences and Engineering Research
Council of Canada, in the form of an operating
grant to E.J.M. andbyConcordia University, by
way of a post-doctoral fellowship to P.C.-F.
References
Blades P.I. & Youngbluth M J . (1980) Morphological,
pliysiological and hchavioural aspeets of mating in
calanoid copepods. Evohuion und Ecology of
Zooplankton Communities (Ed. W. C. Kerfoot). pp. 39-.S1. University Press of New England.
Carter J.C.H.. Dadswctl. M.J.. Roff J.C. & Sprules
W.G. (I'JHIi) Distrihulion and zoogeography of
piiinktonic crustaceans and dipterans in glaciaied
eastern North America. Canaelian Journul of Zoology, 58. 1355-13S7.
107
C'how-Fraser P. | \9Hb) An empirical model to predict
in silu grazing rates of DiaptumiLS minutus Lilljeborg on small algal particles. Canudian Journal
of Fisheries and Aquatic Sciences, 43, 1()65-1()7().
Cfiow-Frascr P. & Wong O.K. {19S6) Dietary change
during development in the frcshwiitcr calanoid
copcpod Epischura lacusiris Forbes. Canadian
Journal of Fisheries and Aquatic Sciences, 43, 938944.
Cole G.A. (1961) Some calnoid copepods trom Arizona with notes on congeneric occurrences of Diaptomus species. Limnology and Oceanographv.
6,432-442.
Davis C C . (1961) Breeding of calanoid copepods in
Lake Erie. Internationale Vereinigunf> fiir Theoretische und Angewandte Limnologie. 14. 933942.
DeFrenza J., Kirner R.J., Maly E.J. & van Lecuwcn
H.C. (1986) The relationships of sex size ratio and
season to mating intensity in some calanoid
copepods. Limnolog\ and Oceanograph\. 31.491496.
Elmore. J.L. (1983) Factt)rs influencing Diapiomus
distributions. An experimental study in subtropical
Florida. Limnology and Oceanographx. 28. 522532.
Fleminger A. (1967) Taxonomy, distribution and
polymorphism in the Labidocera iollae group with
remarks on evolution within the group (Copepoda:
Calanoida). Proceedings of the United States Natural History Museum. 120, 1-61.
Fleminger A. (1975) Geographical distribution and
morphological divergence in American coastalzone planktonic copepods of the genus Labidocera. Estuarine Research (Ed. L. E. Cronin)
Vol. 1. pp. 392-419. Academic Press. New York.
Hopkins C.C.E. ik Machin D. (1977) Patterns of
sperm a lop ho rt' distribution and placement in Euchaeta nurvegica (Copepoda: Calanoida). Journal
of the Marine Biological Association of the United
Kingdom. 57. 113-131.
Hutchinson CJ.E. (1951) Copcpodology for the ornithologist- Ecologw 32. 571-577.
Jacoby C A . & Youngbluth M J . (1983) Mating behaviour in three species of Pseudodiaptomus
(Copepoda: Calanoida). Marine Biology, 76. 7786.
Katona S.K. (1973) Evidence for sex pheromones in
planktonic copepods. Limnology and Oceanography. 18.574-583.
Katona S.K. (1975) Copulation in the copepod Eurytemora afjinis (Poppe). Cru.staceana, 28. 89-94.
Kitlredgc J.S.. "I'akahashi F.T., Lindsey J. & Lasker
R. (1974) Chemical signals in the sea: marine
allelochemics and evolution. Fisheries Bulletin. 11,
1-11.
MalyE.J. (1976) Resource overlap between co-occurring copepods: effects of predation and environmental fluctuation. Canadian Journal of Zoology,
54. 933-940.
Maly E.J. (19K4) Factors affecting distribution and
body size of several Quebec caianoid eopepods.
Crustaceana, Suppl. 7, 315-325.
Rigier F.H. & Langford R.R. (1967) Congeneric oc-
108
Patricia Chow-Fraser and Edward J. Maly
curretices of species of Diaptomus in southern
Ontario lakes. Canadian Journal of Zoology. 45,
81-90.
Sandercock G. (1967) A study of selected mechanisms
for the co-existence of Diaptomus spp. in Clarke
L., Ontario. Limnology and Oceanographv. 12,
97-112.
Sokal R. & Rohlf F.J- (1981) Biometry, 2nd edn.
W. H. Freeman, San Francisco.
Watras C J . (1983) Mate location by diaptomid
copepods. Journal of Plankton Research, 5, 417425.
Zar J.H. (19S4) Bioslalisticat Analysis, 2nd edn. Prentice-Hail, Englewood Cliffs, N.J.
(Manuscript accepted 5 August 1987)